# Pattern Synthesizer

> Extract reusable patterns from examples, signals, or repeated cases. Use when many observations exist and the next need is to identify the stable structure underneath rather than describe each case separately.

- Skill: `alubiama/pattern-synthesizer` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add alubiama/pattern-synthesizer`
- Raw SKILL.md: https://api.skillmd.com/api/skills/alubiama/pattern-synthesizer/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Alubiama (https://skillmd.com/u/alubiama)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/alubiama/pattern-synthesizer

---


# Pattern Synthesizer

## Goal

Find the stable pattern underneath multiple examples or signals.

This skill does not summarize each item. It extracts the reusable structure, recurring mechanism, or common shape that matters for future decisions.

## When To Use

- several cases look similar
- repeated observations need a unifying pattern
- the next step needs a model, not another list

## Scope Boundaries

In scope:

- compare repeated cases
- extract the common mechanism or structure
- note where the pattern breaks

Out of scope by default:

- statistical proof of the pattern
- forcing pattern where only coincidence exists

## Workflow

1. List the examples or signals.
2. Compare common structure.
3. Extract the repeating mechanism.
4. Note exceptions or breakpoints.
5. State the reusable pattern and why it matters.

## Output Contract

1. `input_examples`
2. `pattern_candidate`
3. `supporting_structure`
4. `exceptions`
5. `why_the_pattern_matters`
6. `recommended_followup`
7. `next_step`

## Guardrails

- Do not force one pattern across clearly different cases.
- Keep exceptions visible.
- Prefer a humble pattern over an overfit grand theory.